A refrigerator door recognition method and system based on deep convolutional neural network

Abstract

The invention discloses a refrigerator door recognition method and system based on a deep convolutional neural network, wherein the method includes collecting images of all models of refrigerator doors that require secondary verification on the production line; Model classification, data cleaning and data expansion preprocessing; extract the texture features of the preprocessed refrigerator door image, and construct a deep convolutional neural network structure; use the preprocessed refrigerator door image as training data, use the depth convolution Convolutional neural network structure to train the refrigerator door recognition model; use the trained refrigerator door recognition model to test the accuracy of the refrigerator door image to be recognized, if the test accuracy does not meet the industrial application standard, rebuild the deep convolutional neural network structure To train the refrigerator door recognition model until it reaches the industrial application standard.

Description

technical field [0001] The invention belongs to the field of image classification and recognition, and in particular relates to a refrigerator door recognition method and system based on a deep convolutional neural network. Background technique [0002] In recent years, deep learning techniques, especially convolutional neural networks, have been widely used in image recognition tasks such as image classification, object detection, and image segmentation. At the same time, with the progress of society and the continuous deepening of industrial automation, the use of computer-aided industrial production can save manpower and production costs on the one hand, and improve production efficiency on the other hand. [0003] On the refrigerator production line in the factory, the finished refrigerator needs to undergo secondary verification before packing to ensure that the color and appearance of the refrigerator door are consistent with the model of the refrigerator produced. At...

AI Technical Summary

Problems solved by technology

At present, factories mainly use workers to carry out secondary verification of refrigerator models on the production line, which requires workers not only to memorize the color and appearance of doors corresponding to a large number of refrigerator models in advance, but also to have a certain degree of proficiency and business level

In addition, in the process of secondary verification of refrigerator models, on the one hand, the work efficiency of workers will affect production efficiency. When there is a new refrigerator model that needs to be verified and identified, workers need to spend a lot of time learning and memorizing; on the other hand, the production cost and expenditures also lead to waste

[0004] To sum up, in the prior art, there is still no effective solution to the problem of slow and high cost of refrigerator door recognition

Images